AN INTRODUCTION TO BOOLEAN ALGEBRAS

Transcription

1 California State University, San Bernardino CSUSB ScholarWorks Electronic Theses, Projects, and Dissertations Office of Graduate Studies AN INTRODUCTION TO BOOLEAN ALGEBRAS Amy Schardijn California State University - San Bernardino, schardia@coyote.csusb.edu Follow this and additional works at: Part of the Algebra Commons Recommended Citation Schardijn, Amy, "AN INTRODUCTION TO BOOLEAN ALGEBRAS" (2016). Electronic Theses, Projects, and Dissertations. Paper 421. This Thesis is brought to you for free and open access by the Office of Graduate Studies at CSUSB ScholarWorks. It has been accepted for inclusion in Electronic Theses, Projects, and Dissertations by an authorized administrator of CSUSB ScholarWorks. For more information, please contact scholarworks@csusb.edu.

2 An Introduction to Boolean Algebras A Thesis Presented to the Faculty of California State University, San Bernardino In Partial Fulfillment of the Requirements for the Degree Master of Arts in Mathematics by Amy Michiel Schardijn December 2016

3 An Introduction to Boolean Algebras A Thesis Presented to the Faculty of California State University, San Bernardino by Amy Michiel Schardijn December 2016 Approved by: Dr. Giovanna Llosent, Committee Chair Date Dr. Jeremy Aikin, Committee Member Dr. Corey Dunn, Committee Member Dr. Charles Stanton, Chair, Department of Mathematics Dr. Corey Dunn Graduate Coordinator, Department of Mathematics

4 iii Abstract This thesis discusses the topic of Boolean algebras. In order to build intuitive understanding of the topic, research began with the investigation of Boolean algebras in the area of Abstract Algebra. The content of this initial research used a particular notation. The ideas of partially ordered sets, lattices, least upper bounds, and greatest lower bounds were used to define the structure of a Boolean algebra. From this fundamental understanding, we were able to study atoms, Boolean algebra isomorphisms, and Stone s Representation Theorem for finite Boolean algebras. We also verified and proved many properties involving Boolean algebras and related structures. We then expanded our study to more thoroughly developed theory. This comprehensive theory was more abstract and required the use of a different, more universal, notation. We continued examining least upper and greatest lower bounds but extended our knowledge to subalgebras and families of subsets. The notions of cardinality, cellularity, and pairwise disjoint families were investigated, defined, and then used to understand the Erdös-Tarski Theorem. Lastly, this study concluded with the investigation of denseness and incomparability as well as normal forms and the completion of Boolean algebras.

5 iv Acknowledgements I would like to thank my advisor, Dr. Giovanna Llosent. You have unwaveringly supported me through this journey. You have given me much needed advice and guidance and always remained patient with me. I would not have completed this achievement without you. Though I can say thank you, I will never be able to truly express how appreciative I am for your impact on my life. Thank you for everything. I would like to thank my committee members, Dr. Jeremy Aikin and Dr. Corey Dunn. Dr. Aikin, thank you for agreeing to be on my committee and for devoting your time to help me on top of all of the other students you so generously support. Dr. Dunn, thank you for your support and advice over these past years. You have helped me immensely with making the best decisions for my academic career. I also wish thank Joyce Ahlgren. Without ever hesitating, you always help me in any and every way that you possibly can. I am very thankful for you. To my family, you always believe in me, never doubting that I can accomplish anything I set out to achieve. Tunji, you believe in me more than I believe in myself most of the time. Whenever I am struggling, I turn to you and to God and find renewed strength. You encourage me, you help me grow, and you inspire me to think harder and dream bigger each day. Thank you for taking this journey with me. To my high school counselor, Mr. Rodney Hoopai, without whom I would not be who or where I am today. You asked if I planned to go to college. My response was, I guess... Then you sat with me in your office and we filled out the application together. Thank you.

6 v Table of Contents Abstract Acknowledgements List of Figures iii iv vi 1 Introduction Notation Preliminaries Partially Ordered Sets and Lattices 6 3 Introduction to Boolean Algebras 18 4 Atoms and Ultrafilters 32 5 More on Boolean Algebras 36 6 Conclusion 45 Bibliography 46

7 vi List of Figures 2.1 Hasse Diagram of S Hasse Diagram of the Lattice of Subgroups of S Hasse Diagram of the Lattice of Subgroups of D Hasse Diagram of the 2-element Boolean Algebra Hasse Diagram of a 4-element Boolean Algebra Hasse Diagram of an 8-element Boolean Algebra

8 1 Chapter 1 Introduction The topic of Boolean algebra is a branch of algebra first introduced by George Boole that involves mathematical logic. Computations with Boolean algebras involve the operations of greatest lower bound, least upper bound, and complementation. The structure of a Boolean algebra develops from a partially ordered set and, subsequently, from a lattice. The literature on this subject uses two different types of notation. Because of this fact, this thesis contains multiple notations to denote a single concept. For example, and both denote least upper bound. The varied notation will allow us to understand all of the necessary definitions and it will also allow the reader to be familiar with the different notation when reading various articles on the subject. The motivation to study Boolean algebras comes from an interest in set theory and mathematical logic, as well as a desire understand some of the mathematics that apply to the computer design process.

9 2 1.1 Notation N {0, 1, 2, 3,... } The set of natural numbers Z {0 ± 1, ±2, ±3,... } The set of integers Z Z { \ {0} Q a b a Z, b Z } The set of rational numbers Q Q \ {0} R The set of real numbers R R { \ {0} } C a + bi a, b R, i = 1 The set of complex numbers C C \ {0} is an element of The union of sets A B = {x x A or x B} The intersection of sets A B = {x x A and x B} or { } The empty set Subset Proper subset gcd(a, d) The greatest common divisor of a and b (a, b Z) lcm(a, b) The least common multiple of a and b (a, b Z) min(a, b) The minimum or least value of a and b max(a, b) The maximum or greatest value a and b a < b a b and a b ω The set of all finite ordinals

10 3 1.2 Preliminaries The following will help provide the necessary background for the content discussed in this thesis. Definition 1.1. A set S is a collection of elements. A subset A of a set S is a set in which all of the elements of A are elements of S. For s S, the complement of s is all of S except s, denoted S \ s. A set is considered to be open if its complement is closed and closed if its complement is open. Definition 1.2. A number a is congruent modulo n, or mod n, to b if and only if n divides (b a). We write a b (mod n). Definition 1.3. For a set S, a partition is a collection of nonempty disjoint subsets of S whose union is the set S. Definition 1.4. A binary operation on a set G is a function that, for each pair of elements in G, assigns a single element in G. Definition 1.5. When a binary operation, denoted is applied to a nonempty set G, then G is a group under if the following properties are true: (i) For any three elements a, b, c G, then (a b) c = a (b c). This is called the associative property. (ii) There exists an element e G such that, for all a G, a e = a and e a = a. The element e is called the identity. (iii) For each a G, there exists an inverse of a, denoted a 1, in G such that a a 1 = e and a 1 a = e. Definition 1.6. The number of elements of a group G is called the order of G and is denoted G. Definition 1.7. An ordinal number, or ordinal, is the ordering type of a well ordered set. The cardinal number, or cardinality of a set X, is the least ordinal with which X has a one-to-one correspondence. The cardinality of X is denoted X. Definition 1.8. A set X is finite if it contains a finite number of elements. A subset A of X is cofinite in X if X \ A is finite. A group G is finite if the order of G, G, is finite.

11 4 Definition 1.9. If G is a group and a G, then the order of a, denoted a, is the smallest number n (where n Z ), such that a n = e (where e is the identity element of G). Note: if the operation on G is addition, then a = n such that na = e = 0. Definition If A and B are sets, then a map (or function) φ from A to B, denoted φ : A B, assigns to each element of A exactly one element from B. We use the notation φ : a b for a A, b B such that φ(a) = b. If φ : A B, then the image of an element a A is φ(a) B, where φ : a φ(a). Then we call a the preimage of φ(a). Definition A map φ : A B is well defined if it unambiguously assigns one and only one element of B to each element of A; that is, if x = y, for x, y A, then φ(x) = φ(y). Definition A map φ : A B is one to one (or an injection) if whenever a 1 a 2, then φ(a 1 ) φ(a 2 ). Also, φ : A B is onto (or a surjection) if for every b B, there is an element a A such that φ(a) = b. If φ : A B is one to one and onto, then we say it is a bijection. Definition If G is a group with operation and H is a group with operation, then the map φ : G H is a group homomorphism if φ(a b) = φ(a) φ(b) for all a, b G. Definition If φ : G H is a homomorphism and e is the identity element in H, then the kernel of φ, denoted Kerφ is the set {g G φ(g) = e}. Definition A homomorphism φ : G H is an group isomorphism if φ is a bijection and we say that G and H are isomorphic, denoted G = H. Definition If φ : A B and there exists a map φ 1 : B A such that φ 1 φ is the identity map on A (that is, φ 1 φ : A A maps a to φ 1 φ(a) for all a A), and φ φ 1 is the identity map on B, then φ : A B is said to be invertible. In this case, φ 1 is the inverse of φ. Definition A topological space is a nonempty set X with a family of subsets, that are all open sets, such that and X are open sets, the union of an any number of open sets is open, and the intersection of a finite amount of open sets is open.

12 5 Definition For a topological space X, a cover C of X is a collection of sets M i, where i I, such that X = i I M i. If M i are open sets, then C is an open cover. A subcover of C is a subset of C that is also a cover of X. Definition A topological space X is compact if every open cover of X has a finite subcover. Definition A topological space X is a Hausdorff space if, for any x, y X with x y, there exist open sets M and N, with x M and y N, such that M N = { }. Axiom 1.1. Every nonempty set can be well ordered.

13 6 Chapter 2 Partially Ordered Sets and Lattices In order to study Boolean algebras, we begin by developing an understanding of partially ordered sets, followed by lattices. A lattice, equipped with additional characteristics, is the underlying structure of a Boolean algebra. Definition 2.1. If S is a set and is a relation on S, then S is a partially ordered set, denoted S,, if the following axioms are true for any a, b, c S: (i) a a (reflexivity). (ii) If a b and b a, then a = b (antisymmetry). (iii) If a b and b c, then a c (transitivity). Definition 2.2. An element b S, is said to cover an element a S if a < b and there is no c S such that a < c < b. Definition 2.3. A partially ordered set S, that also satisfies a b or b a for any a, b S (comparability) is called a linearly ordered set or a total order set. A totally ordered set is well ordred if and only if all of its nonempty subsets contain a least element. Example 2.1. The set of integers, Z, is a partially ordered set, where a b is defined as a less than or equal to b. Exercise 2.1. Determine whether the indicated set or relation is an example of a partially ordered set: the set of integers Z with a b to mean a b (a divides b).

14 7 Solution. a b = a m = b for m Z. Then a a (reflexivity) since a m = a, where m = 1 Z. Let a b and b a for a, b Z. Then a m = b and b n = a for m, n Z. So a m = b = b n m = b (since a = b n) = b (n m) = b. Similarly, b n = a = a m n = a (since b = a m) = a (m n) = a. So b (m n) = b = m n = b b = 1 and a (m n) = a = m n = a a = 1. But m, n Z, so m = ±1 and n = ±1 since m n = 1. Then a m = b = a (±1) = b = a = ±b. Then a b (when a = b) and antisymmetry does not hold. Therefore, this is not an example of a partially ordered set. Exercise 2.2. Determine whether the indicated set or relation is an example of a partially ordered set: the set of natural numbers N with a b to mean a b (a divides b). Solution. a b = a m = b for m N. Then a a (reflexivity) since a m = a, where m = 1 N. Let a b and b a for a, b N. Then a m = b and b n = a for m, n N. So a m n = a since b = a m and b n m = b since a = b n. Then a (m n) = a = (m n) = 1 and b (n m) = b = n m = 1. Then m = 1 and n = 1 since m n = 1 for m, n N. So a m = b n (since a m = b and b n = a) and a 1 = b 1. Then a = b (antisymmetry). Now, let c N and a b and b c. Then a m = b and b n = c. So a m n = c, but (m n) N since the set of natural numbers is closed under multiplication. Then a (m n) = c = a c (transitivity). Therefore, the set of natural numbers N with a b to mean a divides b is an example of a partially ordered set. Definition 2.4. For a set A, a subset B of A is cofinal if for every a A there exists b B such that a b. If A is a partially ordered set, then the smallest of the cardinalities of the cofinal subsets of A is called the cofinality of A, denoted cf(a). Definition 2.5. A cardinal number k is regular if and only if k is infinite and the cofinality of k is k. If k is an infinite cardinal that is not regular, then k is singular. Definition 2.6. A nonempty set H of a group G is a subgroup of G if H is a subset of G and it is a group under the same operation as G. It is denoted by H G. Definition 2.7. A subgroup H of a group G is maximal if there does not exist a subgroup J of G such that H < J < G.

15 8 Definition 2.8. If G is a group and a G, then a = {a n n Z} is the cyclic subgroup generated by a. Note: if the operation on G is addition, then a = {na n Z}. Definition 2.9. If H is a subgroup of G and a G, then the set ah = {ah h H} is a left coset of H in G and the set Ha = {ha h H} is a right coset of H in G. Definition A subgroup H of G is a normal subgroup of G if for all a G, then ah = Ha; or equivalently, if for all a G, then aha 1 = H. We denote H as a normal subgroup of G by H G. Definition If H is a normal subgroup of G, then the set of cosets of H in G, denoted G/H = {ah a G} and read G modulo H, is a group under the operation (ah)(bh) = (ab)h and is called the quotient group (or factor group) of G by H. Exercise 2.3. Determine whether the indicated set or relation is an example of a partially ordered set: the set of all subgroups of a group G with H K to mean that H is a normal subgroup of K, denoted H K. Solution. H H (reflexivity) since for all h H, then hh = Hh. That is, {hh h H} = {hh h H}. Now consider H K and K H. If H K, then H K. Similarly, if K H, then K H. Thererfore H = K (antisymmetry). For transitivity, we would show that for H K and K L, then H L. Consider the following sets: L = A 4, K = {e, (12)(34), (13)(24), (14)(23)}, and H = {e, (12)(34)}. Then H K and K L. But for h = (12)(34) and l = (123), then H L since (14)(23) / H. Therefore, the set of all subgroups of a group G with H K to mean that H is a normal subgroup of K is not an example of a partially ordered set. Definition A group G with operation is Abelian if, for all a, b G, a b = b a. This property is called commutativity. Definition A ring is a set R equipped with two operations (, +) such that, for all a, b, c R, then: (i) R is an Abelian group under + (ii) a b R (closure)

16 9 (iii) a (b c) = (a b) c (associativity) (iv) a (b + c) = a b + a c and (b + c) a = b a + c a (distributivity) Definition A ring R is a commutative ring if, for all a, b R, then a b = b a. It is a commutative ring with unity if there exits 1 R such that a 1 = a = 1 a for all a R. Definition A subring S of a ring R is a subset S of R such that S is a ring under the operations of R. Definition An ideal of a ring R is a subring S such that for every r R and every s S, then rs S and sr S. Exercise 2.4. Determine whether the indicated set or relation is an example of a partially ordered set: the set of all ideals in a ring R with I J to mean that I is an ideal in J. Solution. Since I is an ideal in J, then for all b J and for all a I, ba I and ab I. Also, I is a subring. J is an ideal in R, which implies that for all r R and for all b J, then rb J and br J. Also, J is a subring. Now I I (reflexivity) since for all a I, then aa I (since I is a subring). Let I J and J I. Then I is a subring of J and J is a subring of I. Therefore I = J (antisymmetry). Consider I J and J K. We want to show that I K. Consider the following: let R be the ring of polynomials of x over Q. Then K = a n x n a 5 x 5 + a 4 x 4 + a 3 x 3 + a 2 x 2 + a 1 x, J = a n x n a 5 x 5 + a 4 x 4 + a 3 x 3 + a 2 x 2, and I = a n x n a 5 x 5 + a 4 x 4 + a 2 x 2 are ideals in R such that I J and J K. But I is not an ideal in K since x 2 I and x K but x x 2 = x 3 / I. Therefore transitivity fails. So the set of all ideals in a ring R with I J to mean that I is an ideal in J is not an example of a partially ordered set. Definition A nonzero element a in a commutative ring R is a zero-divisor if there is a nonzero element b R such that a b = 0. Definition A commutative ring with unity and no zero-divisors is an integral domain.

17 10 Definition In a commutative ring with unity R, an element a in R is a unit if there exists an inverse element of a in R, denoted a 1, such that a a 1 = 1 = a 1 a. Definition A field F is a commutative ring with unity such that every nonzero element in F is a unit. Definition For a commutative ring R with identity I R, a ring A with identity I A, and a ring homomorphism f : R A such that f(i R ) = I A, an R-algebra is the ring A together with f such that f(r), is contained in the center of A (where the center of A is the set of all elements a A such that a commutes with every element in A). Definition If S is a set, then the power set of S, denoted P (S), is the set of all subsets of S, where addition and multiplication on P (S) are defined as A + B = {c c A B, c / A B} and A B = {c c A B}. Definition A Hasse diagram displays the elements of a partially ordered set S, such that a line is drawn upward from an element a to another element b if b covers a. Definition If A is a subset of S, where S, is a partially ordered set, then u S is an upper bound for A if for all a A, then a u. If v S and v a for all a A, then v is a lower bound for A. The element w S is a least upper bound (lub) of A if it is an upper bound of A and w u for each upper bound u of A. The element w S is a greatest lower bound (glb) of A if it is a lower bound of A and v w for each lower bound v of A. The lub of a set is also called the supremum, denoted sup, and the glb is also called the infimum, denoted inf. Proposition 2.1. If S, is a partially ordered set and A is a subset of S, then (i) the lub of A is unique, if it exists, and (ii) the glb of A is unique, if it exists. Proof. (i) Suppose w 1 and w 2 are each a lub of A. Then w 1 and w 2 are each an upper bound of A. Now, since w 1 is a least upper bound of A, then w 1 u, for all upper bounds u of A. So w 1 w 2. Similarly, since w 2 is a least upper bound of A, then w 2 u, and so, w 2 w 1. Then, by antisymmetry of partially ordered sets, w 1 = w 2 and therefore, any lub of A is unique.

18 11 (ii) Suppose w 1 and w 2 are each a glb of A. Then w 1 and w 2 are each a lower bound of A. Since w 1 is a greatest lower bound of A, then w 2 w 1. Similarly, since w 2 is a greatest lower bound of A, then w 1 w 2. Then, by antisymmetry, w 1 = w 2, and therefore any glb of A is unique. Example 2.2. Let S = {a, b, c}. Then the power set of S is P (S) = {{ }, {a}, {b}, {c}, {a, b}, {a, c}, {b, c}, {a, b, c}}. The Hasse diagram of S is {a, b, c} {a, b} {a, c} {b, c} {a} {b} {c} { } Figure 2.1: Hasse Diagram of S {a}. Let X S such that X = {{a}, {a, b}}. Then lub{x} = {a, b} and glb{x} = Definition A lattice is a partially ordered set L, such that for any pair of elements a, b L: (i) The lub of {a, b}, denoted a b, exists. a b is read as a join b or a or b. (ii) The glb of {a, b}, denoted a b, exists. a b is read as a meet b or a and b. Alternatively, lub{x, y} = x + y and glb{x, y} = x y. A lattice is complete if each subset of L, has a lub and a glb. Definition A subgroup lattice is a lattice whose elements are the subgroups of a group and whose relation is set inclusion. Definition If L, is a lattice, then (i) An element 0 L is called a zero if 0 a for all a L.

19 12 (ii) An element 1 L is called a unity if a 1 for all a L. (iii) An element a L is called the complement of a if for any a L, then a a = 1 and a a = 0. Alternatively, if x is the complement of x, then x + ( x) = 1 and x ( x) = 0. Example 2.3. Let S = {1, 2, 3, 5, 6, 10, 15, 30} be the set of positive divisors of 30. Then S is a lattice with unity = 30 and zero = 1. Exercise 2.5. Determine whether the indicated set or relation is an example of a lattice: the set of all positive divisors of 70 with a b to mean a divides b. Solution. Let S = {1, 2, 5, 7, 10, 14, 35, 70}. If a b, then b = a n for n Z. First, a a since a = a 1 (reflexivity). Now suppose a b and b a. Then b = a n and a = b m for m, n Z. So b = a n = (b m) n = b (m n) = m n = 1 = m = 1, n = 1 (since a > 0, b > 0). Then a = b (antisymmetry). Suppose a b and b c. Then b = a n and c = b m for m, n Z. So c = b m = (a n) m = a (n m) (since Z is closed under multiplication). Therefore c = a k for some k = n m Z and so a c (transitivity). For any a, b S, lcm(a, b) S and gcd(a, b) S. So a b = lcm(a, b) (lub exists) and a b = gcd(a, b) (glb exists). Therefore S is an example of a lattice. Exercise 2.6. Determine whether the indicated set or relation is an example of a lattice: the set of all positive divisors of 60 with a b to mean a divides b. Solution. Let S = {1, 2, 3, 4, 5, 6, 10, 12, 15, 20, 30, 60}. Then a a since a = a 1 (reflexivity). From Exercise 1.5, we can see that for a b and b a, then a = b and for a b and b c, then a c (antisymmetry and transitivity). For any a, b S, then lcm(a, b) S and gcd(a, b) S. So a b = lcm(a, b) and a b = gcd(a, b) (lub and glb exist). Therefore S is an example of a lattice. Definition If A and B are sets, then the Cartesian product of A and B is the collection of ordered pairs of elements of A and B and is denoted A B = {(a, b) a A, b B}. Definition If S is a set and R is a subset of S S, then R is a relation on S. For a, b S, then (a, b) R, or equivalently arb.

20 13 Exercise 2.7. Determine whether the indicated set or relation is an example of a lattice: the set L M = {(a, b) a L, b M} where L and M are lattices, with (a, b) (c, d) to mean that a c L and b d M. Solution. Since L and M are lattices, then a a L and b b M. So (a, b) (a, b) (reflexivity). Suppose (a, b) (c, d) and (c, d) (a, b). Then a c L, b d M and c a L, d b M. Since L and M are lattices, a c L and c a L = a = c. Also, b d M and d b M = b = d. Therefore (a, b) = (c, d) (antisymmetry). Now consider (a, b) (c, d) and (c, d) (e, f). Then a c and c e L. Also, b d and d f M. Since L and M are lattices, a c, c e L = a e and b d, d f M = b f. Therefore (a, b) (e, f) (transitivity). For any a, c L, a c L and a c L since L is a lattice. Similarly, for any b, d M, b d M and b d M since M is a lattice. Then (a, b) (c, d) = (a c, b d) L M and (a, b) (c, d) = (a c, b d) L M (lub and glb exist). Therefore, L M is an example of a lattice. Definition If R is a relation on a set S and a, b, c S then R is an equivalence relation on S if the following properties are true: (i) For all a S, (a, a) R. This is called reflexivity. (ii) If (a, b) R, then (b, a) R. This is called symmetry. (iii) If (a, b) R and (b, c) R, then (a, c) R. This is called transitivity. Definition If R is an equivalence relation on a set S and a S, then the set [a] = {b S (a, b) R} is the equivalence class of S containing a. Theorem 2.1. If L, is a lattice and a, b, c L,, then (i) a b = b a and a b = b a (Commutativity) (ii) a (b c) = (a b) c and a (b c) = (a b) c (Associativity) (iii) a a = a and a a = a (Idempotence) (iv) (a b) a = a and (a b) a = a (Absorption) Proof. (i) a b = lub{a, b} and b a = lub{b, a}. Since {a, b} = {b, a} as sets, then a b = b a. Similarly, a b = glb{a, b} and b a = glb{b, a}. Since {a, b} = {b, a} as sets, then a b = b a.

21 14 (ii) a (b c) = lub{a, lub{b, c}} and (a b) c = lub{lub{a, b}, c}. Suppose lub{a, b, c} = d. Then a (b c) = lub{a, lub{b, c}} = lub{a, d} = d = lub{d, c} = lub{lub{a, b}, c} = (a b) c. Therefore, a (b c) = (a b) c). Similarly, a (b c) = glb{a, glb{b, c}} and (a b) c = glb{glb{a, b}, c} and therefore, a (b c) = (a b) c. (iii) a a = lub{a, a} = a, and therefore a a = a. Also, a a = glb{a, a} = a, and therefore a a = a. (iv) We know that a a b and a a by definition of lub and by reflexivity. Then a a (a b) a. Then since a a = a (idempotence), we have a (a b) a. But (a b) a a by definition of glb. Therefore, by antisymmetry, (a b) a = a. Similarly, we know that a b a and a a (glb and reflexivity). So (a b) a a a = a (idempotence). But a (a b) a (lub). Therefore, (a b) a = a by antisymmetry. Proposition 2.2. Any statement that is true for every lattice continues to be true if (i) and are interchanged throughout the statement and (ii) and are interchanged throughout the statement. The new statement that is obtained from interchanging and and also and is called the dual of the original statement. A dual statement for a Boolean Algebra is obtained by interchanging + and and interchanging 0 and 1. If a statement is true for every Boolean algebra, then so is its dual statement. Proof. Let φ represent a statement that holds in every Boolean algebra. Consider a Boolean algebra B and let B be the dual algebra of B, that is, the Boolean algebra resulting from interchanging +,, 0, and 1 in B for, +, 1, and 0, respectively, in B. Let φ be the dual of φ. If φ holds in any Boolean algebra, say B, then by interchanging + and, also 0 and 1, we have that φ holds in B. Theorem 2.2. If L is a set with operations and such that the axioms of commutativity, associativity, idempotence, and absorption are satisfied, and on L is defined by a b if and only if a b = b or a b = a (for a, b L), then L, is a lattice. Proof. We begin by showing that a b = b and a b = a are equivalent. Consider a b = b. Then a b = a (a b) = (a b) a = a (by commutativity and absorption). So a b = b implies a b = a. Conversely, if a b = a, then a b = (a b) b = (b a) b = b. And

22 15 so a b = a implies a b = b. Now we want to show that L is a partially ordered set under. (i) (Reflexivity) If a L, then a a = a (idempotence). Then a a (since a b if and only if a b = a). (ii) (Antisymmetry) If a b then a b = a. If b a, then b a = b. Since commutativity is satisfied under, then a b = b a and therefore a = b. (iii) (Transitivity) If a b, then a b = a. If b c, then b c = b. Since associativity is satisfied under, a c = (a b) c = a (b c) = a b = a. Therefore a c = a, and so a c. Lastly, we want to show that a b is the lub{a, b} and a b is the glb{a, b}. Let us evaluate a b. Consider a (a b) = (a b) a = a. Then a a b. Also, b (a b) = (a b) b = (b a) b = b. Then b a b. Therefore, a b is an upper bound for a and b. Now, let c L such that a c and b c. Then a c = c and b c = c. So (a b) c = a (b c) = a c = c. Therefore a b c, so a b is the lub{a, b}. By duality (interchanging and and also and ), then glb{a, b} = a b. Exercise 2.8. Show that the lattice L is distributive if and only if for all a, b, c L, then a b = a c and a b = a c = b = c. Solution. ( = ) Let L be a distributive lattice. We want to show that for all a, b, c L, if a b = a c and a b = a c, then b = c. Since L is distributive, then we have the following: b = (a b) b (absorption) = (a c) b (since a b = a c) = b (a c) (commutativity) = (b a) (b c) (distribution) = (a b) (c b) (commutativity) = (a c) (c b) (since a b = a c) = (c a) (c b) (commutativity) = c (a b) (distribution) = c (a c) (since a b = a c) = c.

23 16 Therefore, b = c. ( = ) Now assume that for a, b, c in a lattice L, if a b = a c and a b = a c = b = c. We want to show that L is distributive. Consider (a b) (a c). (a b) (a c) = (a c) (a c) (since a b = a c) = a c (idempotence) = a (c c) (idempotence) = a (b b) (since b = c). Similarly, (a b) (a c) = (a c) (a c) = a c = a (c c) = a (b c). Therefore, L is distributive. Exercise 2.9. Construct the Hasse diagram for the lattice of subgroups of S 3. Solution. The elements of S 3 are e, (12), (13), (23), (123), and (132). The subgroups of S 3 are { e, (12)}, {e, (13)}, {e, (23)}, {e, (123), (132)}. S 3 {e, (12)} {e, (13)} {e, (23)} {e, (123), (132)} { e} Figure 2.2: Hasse Diagram of the Lattice of Subgroups of S 3 Exercise Construct the Hasse diagram for the lattice of subgroups of D 4.

24 17 Solution. D 4 = {ρ 0, ρ, ρ 2, ρ 3, τ, ρτ, ρ 2 τ, ρ 3 τ}. The subgroups of D 4 are generated by the elements in the Hasse diagram: D 4 e, ρ 2, τ, ρ 2 τ e, ρ, ρ 2, ρ 3 e, ρ 2, ρτ, ρ 3 τ e, τ e, ρ 2 τ e, ρ 2 e, ρτ e, ρ 3 τ e Figure 2.3: Hasse Diagram of the Lattice of Subgroups of D 4

25 18 Chapter 3 Introduction to Boolean Algebras Now that we ve established the notion of lattices, we can begin exploring Boolean algebras and some of their characteristics. Definition 3.1. If B is a lattice with zero and unity, then B is a Boolean algebra if (i) For all a, b, c B, a (b c) = (a b) (a c) and a (b c) = (a b) (a c) (Distributivity) (ii) For each a B, there exists an a B such that a a = 1 and a a = 0 (Complements). Alternatively, a Boolean Algebra is a structure (A, +,,, 0, 1) with binary operations + and, unary operation, zero element 0, and unity element 1, such that for all x, y, z A (i) x + (y + z) = (x + y) + z and x (y z) = (x y) z (associativity) (ii) x + y = y + x and x y = y x (commutativity) (iii) x + (x y) = x and x (x + y) = x (absorption) (iv) x (y + z) = (x y) + (x z) and x + (y z) = (x + y) (x + z) (distributivity) (v) x + ( x) = 1 and x ( x) = 0 (complementation). Note: The following operations are equivalent: and, and +, and. Example 3.1. S = {1, 2, 3, 5, 6, 10, 15, 30} is a Boolean algebra such that, for a, b S, then a b means a divides b. Then a b = lcm{a, b} and a b = gcf{a, b}. Definition 3.2. If B is a Boolean algebra with operations + B, B, B and zero and unity,

26 19 0 B and 1 B, then A is a subalgebra of B if A B and 0 A = 0 B, 1 A = 1 B, and + A, A, A are equivalent to the restrictions of + B, B, B to A. Definition 3.3. If A is a set and P (A) is its power set, then P (A) is the power set algebra of A equipped with zero, unity A, and the operations,, and, where a = A \ a (the complement of a). Definition 3.4. For a set X = {x}, the power set algebra P (X) is {0, 1}, where 0 = { } and 1 = X, is called the two-element algebra and is denoted 2. Definition 3.5. For a set A, a collection of subsets M of A is an algebra of sets if M, for M 1, M 2 M then M 1 M 2 M and M 1 M 2 M, and A \ M 1 M for M 1 M. Equivalently, if M is a subalgebra of the power set algebra P (A), then M is an algebra of sets over A. If a Boolean algebra B is an algebra of sets over A, then B is an algebra of sets. Definition 3.6. A finite-cofinite algebra A on X is the algebra of sets over X such that A is the set of all a X, where a is either finite or cofinite. Definition 3.7. For a subset X of a Boolean algebra B, the subalgebra generated by X in B is X = {A B X A and A is a subalgebra of B}. If X B and X = B, then X is a set of generators for a Boolean algebra B. Exercise 3.1. Draw the Hasse diagrams of all nonisomorphic Boolean algebras of orders B = 2, 4, or 8. Solution. B = 2 = B = {0, 1} 1 0 Figure 3.1: Hasse Diagram of the 2-element Boolean Algebra

27 20 B = 4 = B = {0, a, b, 1} 1 a b 0 Figure 3.2: Hasse Diagram of a 4-element Boolean Algebra B = 8 = B = {0, a, b, c, d, e, f, 1} = P (S), where S = {x, y, z} {x, y, z} {x, y} {x, z} {y, z} {x} {y} {z} { } Figure 3.3: Hasse Diagram of an 8-element Boolean Algebra Exercise 3.2. Let n N and let B(n) be the set of all positive divisors of n. Show that B(n) with = gcd and = lcm is a Boolean algebra if and only if in the prime factorization of n no prime appears with an exponent 2. Solution. ( = ) Assume B(n) is a Boolean algebra. Suppose x and y are factors of n and y = x. Then x y = 1 and x y = n. But if n has repeated prime factors, then it could be that x y = x. Therefore, in the prime factorization of n no prime appears with an exponent 2.

28 21 ( = ) Assume that in the prime factorization of n, no prime appears with an exponent 2. For x, y B(n), x y = lcm(x, y) and x y = gcd(x, y). We will show that B(n) is a Boolean algebra. Distribution: We know x (y z) = gcd(x, lcm(y, z)). Consider the exponents on the prime factors of x, y, and z. Let x be the product of prime factors with exponents i, y with j, and z with k, where i, j, k can be 0 or 1. Then the exponent of a prime factor in gcd(x, lcm(y, z)) is min(i, max(j, k)), which equals max(min, (i, j), min(i, k)), the exponent of a prime factor in lcm(gcd(x, y), gcd(x, z)) = (x y) (x z). Similarly, x (y z) = (x y) (x z). Complementation: x x = gcd(x, x ) = gcd(x, n x ) = 1 and x x = lcm(x, x ) = lcm(x, n x ) = n. Commutativity: x y = gcd(x, y) = gcd(y, x) = y x and x y = lcm(x, y) = lcm(y, x) = y x. Associativity: x (y z) = lcm(x, lcm(y, z)) = lcm(lcm(x, y), z) = (x y) z and x (y z) = gcd(x, gcd(y, z)) = gcd(gcd(x, y), z) = (x y) z. Absorption: x (x y) = gcd(x, lcm(x, y)) = x and x (x y) = lcm(x, gcd(x, y)) = x. Proposition 3.1. If B is a Boolean algebra, then for any a B, a 1 = a and a 0 = a. Proof. By definition of unity in a lattice, a 1 for all a L. Then a 1 = a. By definition of zero in a lattice, 0 a for all a L. Then a 0 = a. Therefore, 1 is an identity element under and 0 is an identity element under. Proposition 3.2. If B is a Boolean algebra, then (i) the 1 and 0 elements in B are unique, (ii) the complement, a, of a in B is unique. Proof. (i) Let e 1 and e 2 be identity elements in B under and. Then e 1 = e 1 e 2 = e 2 e 1 = e 2. Similarly, e 1 = e 1 e 2 = e 2 e 1 = e 2. Then, by Proposition 3.1, the 1 and 0 elements in B are unique. (ii) Let a in B have complements b and c in B. Then

29 22 b = b 0 (identity) = b (a c) (definition of complement) = (b a) (b c) (distribution) = (a b) (b c) (commutativity) = 1 (b c) (definition of complement) = b c (identity) Therefore b = b c and similarly, c = c b. Then b = b c = c b = c by commutativity. And so, the complement of a is unique. Proposition 3.3. If B is a Boolean algebra, then for any a, b B (i) a 0 = 0 and a 1 = 1 (ii) (a b) = a b and (a b) = a b (de Morgan s Law) (iii) (a ) = a (Involution) (iv) 0 = 1 and 1 = 0. Proof. (i) a 0 = a (a a ) = (a a) a = a a = 0 (by complement, associativity, and idempotence). By duality, a 1 = 1. (ii) We first show that (a b) (a b ) = 0. From (i), we have (a b) (a b ) = [(a b) a ] [(a b) b ] (distributivity) = [(a a ) b] [a (b b )] (associativity and commutativity) = (0 b) (a 0) (complement) = 0 0 = 0 Also, (a b) (a b ) = [a (a b )] [(b (a b )] (distributivity) = [(a a ) b ] [a (b b )] (associativity and commutativity) = (1 b ) (a 1) (complement) = 1 1 = 1 Therefore, a b is the complement of a b and by duality, a b is the complement of a b. (iii) Since a a = a a = 0 and also a a = 1, then a is the complement of a. Now, since complements are unique, then (a ) = a.

30 23 (iv) By definition of complement, 0 0 = 1. Then by commutativity, 0 0 = 1 and so, by Proposition 3.1, 0 = 1. Similarly, 1 1 = 0 by definition. Then 1 1 = 0 and by Proposition 3.1, then 1 = 0. Lemma 3.1. For b, c in a boolean algebra B, the following are equivalent: (i) b c (ii) b c = 0 (iii) b c = 1 Proof. We will show that (i) = (ii), (ii) = (iii), and (iii) = (i), therefore showing that (i), (ii), and (iii) are equivalent. If b c, then b c = c (Theorem 2.2). Then we have Therefore, (i) = (ii). If b c = 0, then b c = b (b c) (b c = c) = b (b c ) (de Morgan s) = (b b ) c (associativity) = 0 c (definition of complement) = 0 Therefore, (ii) = (iii). If b c = 1, then b c = b (c ) (involution) = (b c ) (de Morgan s) = 0 (b c = 0) = 1 (Proposition 3.3) b = b 1 (Proposition 3.1) = b (b c) (b c = 1) = (b b ) (b c) (distribution) = 0 (b c) (definition of complement) = b c (Proposition 3.1) Since b = b c, then by definition of glb, b c. So (iii) = (i), and therefore (i), (ii), and (iii) are equivalent.

31 24 Exercise 3.3. Let B be a Boolean algebra and a, b B. Show that (i) a b if and only if b a (ii) a b if and only if a b = 0 (iii) a b if and only if a b = 1. Solution. (i) ( = ) Suppose a b. Then ( = ) Now suppose b a. Then a b = 0 (Lemma 3.1) = b a = 0 (commutativity) = b (a ) = 0 (complement) = b a (Lemma 3.1) b (a ) = 0 (Lemma 3.1) = b a = 0 (complement) = a b = 0 (commutativity) = a b (Lemma 3.1) (ii) ( = ) Suppose a b. Then a (b ) = 0 = a b = 0 (Lemma 3.1). ( = ) Suppose a b = 0. Then a (b ) = 0 = a b (complement and Lemma 3.1). (iii) ( = ) Suppose a b. Then a b = 1 (Lemma 3.1). ( = ) Suppose a b = 1. Then a b (Lemma 3.1). Definition 3.8. If A and B are algebras and R is a field (or ring), then φ : A B is an algebra homomorphism if, for all r R and a, b A, then (i) φ(ra) = rφ(a) (ii) φ(a + b) = φ(a) + φ(b) (iii) φ(ab) = φ(a)φ(b). If A and B are Boolean algebras and x, y A, then φ is a Boolean algebra homomorphism if (i) φ(0) = 0 and φ(1) = 1 (ii) φ(x + y) = φ(x) + φ(y) (iii) φ(x y) = φ(x) φ(y). (iv) φ( x) = φ(x)

32 25 If φ is one-to-one, it is a monomorphism or embedding of A into B and if φ is onto, it is an endomorphism. Definition 3.9. If A and B are Boolean algebras, then φ : A B is an Boolean algebra isomorphism if (i) φ is bijective (one-to-one and onto) (ii) For all a, b A, then a b A if and only if φ(a) φ(b) B (for this, we say φ is order preserving). Alternately, if φ is a bijective homomorphism, then φ is an isomorphism and A is isomorphic to B, A = B. Definition For a subset p of a Boolean algebra A, the homomorphism from A into the two-element Boolean algebra is called the characteristic homomorphism and is denoted χ p : A 2. Exercise 3.4. Show that if φ : B C is a Boolean algebra isomorphism, then for all a, b B (i) φ(a b) = φ(a) φ(b) (ii) φ(a b) = φ(a) φ(b) Solution. (i) Since φ is a Boolean algebra isomorphism, then φ is one-to-one, onto, and a b if and only if φ(a) φ(b). Consider φ(a) φ(b). Then φ(a) φ(b) = φ(b). Now consider a b. Then a b = b, which implies φ(a b) = φ(b). Therefore, φ(a b) = φ(b) = φ(a) φ(b). (ii) Similar to (i), φ(a) φ(b) = φ(a) φ(b) = φ(a) and a b = a b = a = φ(a b) = φ(a). Therefore, φ(a b) = φ(a) = φ(a) φ(b). Exercise 3.5. Show that if φ : B C is a Boolean algebra isomorphism, then (i) φ(1 B ) = 1 C (ii) φ(0 B ) = 0 C (iii) φ(a ) = (φ(a)) for all a B. Solution.

33 26 (i) For all b B, b 1 B. For all φ(b) C, φ(b) 1 C. Then b 1 B = 1 B. Since φ is an isomorphism, then φ(b 1 B ) = φ(1 B ) and by Exercise 3.4, φ(b) φ(1 B ) = φ(1 B ). Now, by definition of unity, φ(b) 1 C = 1 C in C. Since φ is onto, there exists 1 C1 B such that φ(1 C1 ) = 1 C. If 1 B = 1 C1, then φ(1 B ) = 1 C. Suppose 1 B 1 C1. Then 1 C1 1 B = 1 B (definition of unity) and so, φ(1 C1 1 B ) = φ(1 C1 ) φ(1 B ) = φ(1 B ). But φ(1 C1 ) = 1 C and so φ(1 C1 ) φ(1 B ) = 1 C φ(1 B ) = 1 C. Therefore, φ(1 B ) = 1 C. (ii) For all b B, 0 B b. For all φ(b) C, 0 c φ(b). So 0 B b = 0 B and φ(0 B b) = φ(0 B ) φ(b) = φ(0 B ). In C, 0 C φ(b) = 0 C. Since φ is onto, 0 C1 B such that φ(0 C1 ) = 0 C. If 0 B = 0 C1, then φ(0 B ) = 0 C. Suppose 0 B 0 C1. Then 0 C1 0 B = 0 B (definition of zero) and so, φ(0 C1 0 B ) = φ(0 C1 ) φ(0 B ) = φ(0 B ). But φ(0 C1 ) = 0 C and so φ(0 C1 ) φ(0 B ) = 0 C φ(0 B ) = 0 C. Therefore, φ(0 B ) = 0 C. (iii) By (i) and complementation, 1 C = φ(1 B ) = φ(a a ) = φ(a) φ(a ). But 1 C = φ(a) (φ(a)). So φ(a) φ(a ) = φ(a) (φ(a)). Similary, by (ii) and complementation, 0 C = φ(0 B ) = φ(a a ) = φ(a) φ(a ) and 0 C = φ(a) (φ(a)). Then φ(a) φ(a ) = φ(a) (φ(a)). Therefore, by Exercise 2.8, φ(a ) = (φ(a)). Definition A ring R with unity is a Boolean ring if a a = a and a + a = 0 for all a R. A Boolean ring is commutative. Exercise 3.6. Let R be a Boolean ring with unity 1 and for a, b R define a b = a + b a b and a b = a b. Show that R with and is a Boolean algebra with unity 1, zero element 0, and a = 1 a. Solution. We know that all of the ring axioms hold for R under +,, and, and for all a R, then a a = a and a + a = 0. Also, R is commutative. Complementation: We will show a a = 1 and a a = 0. a a = a + a + ( a) a = a + (1 + ( a)) + ( a) (1 + ( a)) (a = 1 + ( a)) = a + [(1 + ( a)) (1 + ( a))] (distribution) = a + (1 + ( a)) (idempotence) = a + a (a = 1 + ( a)) = 1 ( + complementation)

34 27 Since a a = 0, then a a = a a = 0. Commutativity: We will show a b = b a and a b = b a. a b = a + b + ( a) b = b + a + ( a) b (+ commutativity) = b b + a + ( a) b (idempotence) = b (b + ( a)) + a (distribution and + commutativity) = b + a (absorption) = a + b (+ commutativity) = a (a + ( b)) + b (absorption) = a a + b + ( b) a (distribution and commutativity) = a + b + ( b) a (idempotence) = b + a + ( b) a (+ commutativity) = b a. Since R is commutative, then a b = a b = b a = b a. Absorption: We will show a (a b) = a and a (a b) = a. a (a b) = a + (a b) + ( a) (a b) = a + (a b) + ( a a) b ( associativity) = a + (a b) + 0 b (complementation) = a + (a b) + 0 (glb) = a + (a b) (identity) = a (absorption) a (a b) = a (a + b + ( a) b) = a a + a b + a ( a) b (distribution) = a (a + b) + 0 b (distribution and compl.) = a (a + b) + 0 (glb) = a (a + b) (identity) = a (absorption) Distribution: We will show a (b c) = (a b) (a c) and a (b c) = (a b) (a c). a (b c) = a (b + c + ( b) c) = a (b + c) + a (( b) c) (distribution) = a (b + c) + ( b) (a c) (comm. and assoc.) = a (b + c) ( b) (a c) (identity) = a (b + c) + 0 c + ( b) (a c) (glb) = a (b + c) + ( a a) c + ( b) (a c) (complementation) = a (b + c) + ( a) (a c) + ( b) (a c) (associativity) = (a b) + (a c) + ( a + b) (a c) (distribution) = (a b) + (a c) + ( (a b)) (a c) (de Morgan s) = (a b) (a c)

35 28 a (b c) = a + (b c) + ( a) (b c) = a + (b c) 1 + (b c) ( a) (identity and comm.) = a + [(b c) (1 + ( a))] (distribution) = a + [(b c) (1 + ( a)) (1 + ( a))] (idempotence) = a + [(b (1 + ( a)) (c (1 + ( a))] (comm. and associativity) = a + [(b + b ( a)) (c + c ( a))] (distribution) = a + (b + b ( a)) a + (c + c ( a)) (distribution) = (a + b + ( a) b) (a + c + ( a) c) (associativity and comm.) = (a b) (a c) Associativity: We will show a (b c) = (a b) c and a (b c) = (a b) c. First, we will establish that a + a b = a + b. By distribution, a + a b = (a + a ) (a + b). Then, (a + a ) (a + b) = 1 (a + b) = a + b by complementation and identity. Therefore, a + a b = a + b, which we will refer to as. a (b c) = a + (b + c + ( b) c) + ( a) (b + c + ( b) c) = a + (b + c) + ( a) (b + c) (absorption) = a + (b + c) (1 + ( a)) (distribution) = a + (b + c) a (a = 1 + ( a)) = (a + b) + c ( and + associativity) = (a + b) + c (a + b) ( ) = (a + b) + c (1 + (a + b)) (a = 1 + ( a)) = (a + b) + c + (a + b) c (distribution) = (a + b + ( a) b) + c + (a + b + ( a) b) c (absorption) = (a b) c. Since is associative, a (b c) = a (b c) = (a b) c = (a b) c. Therefore, R is a Boolean algebra. Exercise 3.7. Define, in a Boolean algebra A, the binary operations and by x y = x y and x y = x + y. Prove that 0, 1 and the Boolean operations are definable, in terms of equations, both by and. Solution. We will show that x + y and x y are definable in terms of and. x y = ( x) + ( y) = x + y and x y = ( x) ( y) = x y. Now, we will show x x = 0 and x x = 1. By +, complementation and ( x) = x, then x x = x ( x) = x x = 0 and x x = x + ( x) = x + x = 1.

36 29 Definition For a Boolean algebra B with x, y B, the symmetric difference of x and y is x y = x ( y) + y ( x). Alternately, the symmetric difference of a and b is a + b = (a b ) (a b). Exercise 3.8. Using the following notation, check that br is a Boolean algebra and that brb = B and rbr = R: R = (R,,, 0, 1) is a Boolean ring, B = (B, +,,, 0, 1) is a Boolean algebra, rb = (B,,, 0, 1) where x y = x y + y x, br = (R, +,,, 0, 1) where x + y = x y (x y) and x 1 = x. Solution. We want to show that br is a Boolean algebra. First, we will show that R is commutative. Consider x y. By idempotence and distribution, x y = (x y) (x y) = x x x y y x y y = x x y y x y = x 0 y. Then x y y x = 0. If x = y, then x x x x = x x = 0. So x y = y x and R is commutative. Commutativity: Since R is commutative under and, then x + y = x y (x y) = y x (y x) = y + x. Absorption: We will show x + (x y) = x and x (x + y) = x. x + (x y) = x (x y) (x (x y)) = x (x y) ((x x) y) ( associativity) = x (x y) (x y) (idempotence) = x 0 (x x = 0) = x (x 0 = x). x (x + y) = x (x y (x y)) = x (y (x y) x) ( commutativity) = x y x (x y) x x (distribution) = 0 x (idempotence & x x = 0) = x (x 0 = x). Distribution: We will show x (y + z) = (x y) + (x z) and x + (y z) = (x + y) (x + z). x (y + z) = x (y z (y z)) = (x y) (x z) (x (y z)) (distribution) = (x y) (x z) ((x y) (x z)) (assoc., comm., idemp.) = (x y) + (x z)

37 30 x + (y z) = x (y z) (x (y z)) = (x (y z) x) (x (y z) (y z)) (distribution) = (0 (y z)) (x 0) (comm. & x x = 0) = (y z) x (x x = 0) = (y x) (z x) (idemp., comm., assoc.) = ((x y x) (x y y)) ((x z x) (x z z)) (x x = 0, id, comm.) = (x y (x y)) (x z (x z)) (distribution) = (x + y) (x + z). Complementation: Since complementation holds in R for and and x 0 = x, then x + x = x x (x x) = x x 0 = x x = 1. Associativity: Since R is associative for and, we will show x + (y + z) = (x + y) + z. x + (y + z) = x (y z (y z)) (x (y z (y z)) = (x y z (y z) x) (x y z (y z) y z (y z)) (distribution) = (x x y z (y z)) (x y y z z (y z) (y z)) (commutativity) = (0 y z (y z)) (x 0 0 0) (x x = 0) = (y z (y z)) x (x 0 = x) = (y z y) (y z z) x (distribution) = (0 z) (y 0) (x) (x x = 0, comm.) = (z) (0 y) (x 0) (x 0 = x, comm.) = z (x x y) (x y y) (x x = 0) = z (x y (x y)) (comm. & dist.) = (0 0 0 z) (x y (x y) 0) (x 0 = x) = (x x y y (x y) (x y) z) (x y (x y) z z) (x x = 0) = (x y (x y) z x y (x y)) (x y (x y) z z) (commutativity) = (x y (x y)) z ((x y (x y) z) (distribution) = (x + y) + z. Therefore, br is a Boolean algebra. Now we want to show brb = B and rbr = R. For rbr, we want to show x 1 = x and x y (x y) = x + y and for brb, we want to show x = x 1 and x + y = x y (x y). By complementation and the zero and unity properties, x 1 = (x 1) (1 x) ((x 1) (1 x)) = (x 0) ( x) ((x 0) ( x)) = x. Similarly, x 1 = (x 1) + (1 x) = (x 0) + ( x) = 0 + x = x. x y (x y) = x + y: Now, we will show

38 31 x y (x y) = ((x y + y x) (x y)) + ((x y) (x y + y x)) = ((x y + y x) ( x + y)) + ((x y) ( x + y y + x)) (de Morgan s) = ((x y) ( x + y)) + ((y x) ( x + y)) + ((x y) ((( x + y) y) + (( x + y) x))) (distribution) = (x y x) + (x y y) + (y x x) + (y x y) + ((x y) (( x y) + (y y) + ( x x) + (y x))) (distribution) = (x y) + (y x) + (x y x y) + (x y y x) (comp., zero, dist.) = (x y) + (y x) + (x y) (comp., zero) = x ( y + y) + y x (comm., dist.) = x + (y x) (comp., unity) = x + y x + x (comp., unity) = x + y. Similarly, x y (x y) = ((x y + y x) (x y)) + ((x y) (x y + y x)) = x + y. Therefore, rbr = R and brb = B.

39 32 Chapter 4 Atoms and Ultrafilters We have developed an understanding for Boolean algebras and some of their properties thus far. We can now examine atoms and ultrafilters, which are foundational elements of Boolean algebras. Definition 4.1. If B is a Boolean algebra and a < b, then a B is an atom if 0 < a and there does not exist x B such that 0 < x < a. The set of atoms of B is denoted At B. We say B is atomless if it contains no atoms, and atomic if there exists an atom a x for each positive x B. Example 4.1. For the Boolean algebra S = {1, 2, 3, 5, 6, 10, 15, 30}, the elements 2, 3, and 5 are all atoms in S. Example 4.2. The power set algebra P (X) of a set X and the finite-cofinite algebra on X are both atomic sets. The interval algebra of the real number line and the regular open algebra of the real numbers are both atomless sets. Lemma 4.1. If B is a finite Boolean algebra and b is any nonzero element in B, then there exists an atom a in B such that a b. Proof. If b is an atom in B, then b = a. Suppose b is not an atom in B. Then there exists a 1 B such that 0 < a 1 < b. If a 1 is an atom, then a 1 = a. If a 1 is not an atom, then there exists a 2 B such that

40 33 0 < a 2 < a 1. If a 2 is an atom, then a 2 = a. If a 2 is not an atom, then we continue in this manner and eventually obtain 0 < a n <... < a 2 < a 1 < b since B is finite. Then a n = a. Definition 4.2. A subset p of a Boolean algebra A is a filter if (i) 1 p, (ii) if x, y p, then x y p, and (iii) if x p, y A, and x y, then y p. For a A, p is a principal filter if p = {x X a x}. We say p is the principal filter generated by a. If p = {1}, then p is the trivial filter and if 0 / p, then p is a proper filter. Definition 4.3. If M A for a Boolean algebra A, then M is said to have the finite intersection property if {m 1, m 2,..., m n } for m i M (i = 1, 2,..., n). Definition 4.4. An ultrafilter is a filter p of a Boolean algebra A where x p or x p, but not both, for each x A. If p is proper and x + y p = x p or y p for x, y A, then p is a prime filter. If p is proper and proper filter q of A with p q, then p is a maximal filter. The set of ultrafilters of A is Ult A = {p A p is an ultrafilter of A}. Example 4.3. The set {a A x a}, where A is an algebra of sets over X and x is any point in X, is an ultrafilter. Definition 4.5. The Stone map of a Boolean algebra A is the map s : A P (Ult A), where s(x) = {p Ult A x p}. Definition 4.6. If φ : A B is a Boolean algebra homomorphism and Σ A M exists for M A, then φ preserves Σ A M if Σ B φ[m] exists and φ ( Σ A M ) = Σ B φ[m]. Similarly, φ preserves Π A M if φ ( Π A M ) = Π B φ[m]. For an ultrafilter p of A, p preserves ΣM if, for some m M, ΣM p = m p and preserves ΠM if M p = ΠM p. Exercise 4.1. In a Boolean algebra A, let M A such that ΣM exists. The Stone homomorphism s : A P (Ult A) preserves ΣM if and only if ΣM = ΣM 0 for some finite subset M 0 of M. Similarly, let M and N be subsets of A such that s[m] s[n]. Then there are finite subsets M 0 of M and N 0 of N such that s[m 0 ] s[n 0 ]. Solution.